E§§ IIHNIWIHINHWWUHIHIWlfllllflflfllll'IHHI ---4ESIS This is to certify that the thesis entitled Resistance to Cereal Leaf Beetle, Oulema Melanopus, (L.), of Pubescent Avena Sterilis, Avena Sativa Hybrids presented by Carrie Young has been accepted towards fulfillment of the requirements for M. S. Crop & Soil Sciences KMM . Major professor fiegree in Date August 29, 1979 0-7639 flERWE FINES: 25¢ per day per item RETUMIKS LIBRARY MATERIALS: Place in book return to renew charge from circulation record: RESISTANCE TO CEREAL LEAF BEETLE, OULEMA MELANOPUS, (L.), 0F PUBESCENT AVENA STERILIS, AVENA SATIVA HYBRIDS By Carrie Young A THESIS' Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Crop and Soil Science 1979 ABSTRACT RESISTANCE TO CEREAL LEAF BEETLE, OULEMA MELANOPUS, (L.), 0F PUBESCENT AVENA STERILIS, AVENA SATIVA HYBRIDS By Carrie Young Cereal leaf beetle resistance in wheat is highly corre- lated with length and density of leaf pubescence. This relation- ship also exists in wild Hordeum sp. However, additional criteria for selection for resistance in cats must be used because the level of trichome length and density in oats is much lower than in wheat. The USDA World Collection of Small Grains was screened for field larval feeding damage by USDA-CLB project workers. A pubescent Avena sterilis line, PI 3ll677, was selected because of field re- sistance. Entries C.I. 521, C.I. l625, C.I. 4867, and C.I. 4893 also exhibited some resistance. These five lines, with varieties Froker and Korwood, formed the base populations from which initial testing and resistance studies in oats began. A, satjva entries were not significantly different in oviposition preference tests. Line P.I. 3l1677 was significantly different from C.I. 4867 in number of eggs per plant and exhibited antibiosis in larval weight gain tests. A diallele analysis was made on a series of crosses to ob- serve complementation of genes for resistance. If different genes Carrie Young for resistance located at different loci were present in two lines crosses, transgressive segregation could be expected and increase in host plant resistance to the cereal leaf beetle might result. Susceptible lines were included as checks. Dr. Fred Collins of the University of Arkansas contributed pubescent lines. The pubescence was from A, sterilis entries P.I. 295919 and P.I. 320793. The Arkansas lines and the four most pubescent lines from progeny of the diallele crosses were carried forward for further testing. Segregates within each line were selected on the basis of field larval feeding damage tests, oviposition preference tests and trichome density evaluation. Several of these segregates were more pubescent than the original parents and progeny of the initial cross. Crosses among the superior segregates from these progenies formed the basis for the next cycle of recurrent selection. Within line crosses brought together again genes for resistance from the same parent. The trichome counts of parents, F], and F2 of each line were analyzed to determine the number of loci and dominance or recessiveness of trichome expression. These progeny were tested for host plant-beetle interaction. The F1 exhibited increased pubscence density and decreased egg counts compared to the midparent. In the F2, due to the dif- ficulties with single plant observations, these gains were not confirmed. ACKNOWLEDGEMENTS The author wishes to express her sincere thanks and appreciation to Drs. David H. Smith, Jr., John E. Grafius, and James A. Webster for their suggestions for improvement of research techniques and guidance in preparation of this manuscript. Thanks are extended to Dr. Carter Harrison for reading and criticizing the manuscript. I wish to thank my family and Jeffrey Griffin for their concern, encouragement and emotional support during my studies. ii TABLE OF CONTENTS Page Introduction ........................ 1 Literature Review ..................... 3 Plant Material ....................... 9 Tests ........................... ll Trichome Counts . ................. 13 Resistance as Measured by Oviposition Preference . . l7 Larval Feeding Damage ............... 29 Recurrent Selection Crosses ............ 43 Arkansas-ll Segregates ............... 53 Discussion ......................... 56 Conclusions ........................ 62 List of References ..................... 64 Table 10 ll 12 l3 14 LIST OF TABLES Line Identification Numbers and Pedigree ...... Analysis of Variance of Trichome Counts per Microscope Field of Greenhouse Seed Increased (Fall of 1977) ................. Analysis of Variance of Oviposition Test Egg and Trichome Counts (Spring of T978) ........ Correlation and Regression of Egg Counts on Trichome Number of Oviposition Test (Spring 1978) Analysis of Variance of Oviposition Line Test (Winter of 1979) ................ Correlation and Regression of Egg Count on Trichome Number from Oviposition Preference Test (Winter of T979) ................ Analysis of Variance Using All Thirty Lines from Field Data (Summer of l977) ........... Analysis of Variance of Pubescent Lines from Field Test (Summer of 1977) .............. Duncan's Multiple Range Test of Pubescent Line Means from Field Test (Summer of 1977) ..... Correlation and Regression Between Traits Measured During the Field Test (Summer of l977) . Analysis of Variance and Comparison of Larval Feeding Damage Scores (Summer of 1978) ..... Analysis of Variance of Larval Feeding Damage and "Count" (Summer of l979) ............ Comparison of Means for Larval Feeding Damage and "Count" (Summer of 1979) ............ Analysis of Variance of Trichome Counts per Microscope Field for Eight Lines and Five Planting Dates ................. Page 10 16 23 24 27 28 34 35 36 38 4O 44 45 51 INTRODUCTION The cereal leaf beetle, Oulema melanopus (L.), is a chrysomelidae which was first identified in Michigan in 1962. The insect spread rapidly through Michigan, Indiana and Ohio aided by prevailing westerly winds. The beetle has been identified in Illinois and Wisconsin to the west and far into Canada and several states to the east. This Eurasian insect has long been a pest of small grains in Europe. Wheat, oats and barley are damaged excessively. Many native grasses, rye and maize serve as alternate hosts (55). The United States Department of Agriculture, Michigan State University and Purdue University plant geneticists and entomologists screened the World Collection of Small Grains. Field screening followed by laboratory testing found resistant wheat lines. The more resistant wheat lines were highly pubescent. Moderate resistance was discovered in oats and barley. Further work by Hahn (17, 18) and Lee (22) concluded that oviposition preference, larval feeding damage and plant recovery in barley are independently inherited and that resistance is not due solely to leaf pubescence. Several Avena sativa cereal investigation entries were found to be slightly resistant. An A, sterilis entry, P.I. 311677, was pubescent and exhibited an antibiosis effect on larvae (42). l Cereal investigation entries and P.I. 311677 were used to test the host plant-insect interaction under field and laboratory conditions. The inheritance of oat pubescence and its role in resistance were also studied. LITERATURE REVIEW The cereal leaf beetle, Oulema melanopus (L.), is believed to have been introduced into the United States in the late 1950's. The beetle was first noticed in Berrien county in Southwestern Michigan (4). "Black slugs" were found on plants in small grain fields, and unidentified beetles were also seen. The appearance of these insects prompted spraying of insecticides in 1959 (44). In 1962, the slugs (larvae) and beetles were identified as different life cycle stages of cereal leaf beetle. "Watch for the Cereal Leaf Beetle" (56) was published by the United States Depart- ment of Agriculture in 1963 to alert growers of the presence and life cycle stages of this phytophagous insect. Fearing insect spread to the grain growing areas in the plains, screening of germplasm, insecticide tests and formation of quarantine laws began (29). When infested wheat plants were compared to non-injured plants, a 12% reduction in kernel number and a 22.6% reduction in yield was observed (15). Wilson and Shade (54) found larval in- festation and larval feeding damage highly correlated with yield reduction in oats. Screening of the U.S.D.A. Wheat, Oat and Barley World Collections began in 1962. In the 1963-64 growing season, 34,000 3 entries were tested. All of the oat and barley lines experienced larval feeding damage. Only four percent of the oat lines, Avena .§2-, and two percent of the barley lines, Hordeum sp., had trace damage. Overall, twelve percent of the wheat lines, Triticum sp., had trace or no damage. The plants in the pubescent nursery of the Wheat World Collection had 18% of the entries with trace or no larval feeding damage (11). Those wheat lines with no larval feed- ing damage were highly pubescent (16). Gallun et al. (14) ob- served that not only was wheat less preferred for oviposition than oats and barley, wheat lines had less adult and larval feeding damage. Oat lines showed twice as many eggs as wheat lines. Scanning the origin of resistant lines, no single center of resistant germplasm was discovered. Many of the resistant lines are native to China and Russia. Asia, Asia Minor, Eurasia and Eastern Africa are possible germplasm centers (10). The cereal leaf beetle occurs widely throughout Europe and Asia. The most common and serious outbreaks are in Siberia and the Balkans (44). Re- sistance in varieties where the beetle is not indigenous is due to inadvertent use of resistant parents in the breeding program (10). Resistance may be due to the plants' physiological state of development, type of vegetative growth or disease susceptibility (36). When many lines are planted together, individual lines may miss infestation and appear resistant by chance alone. Vernaliza- tion of winter types causes morphological and physiological changes which may affect host selection of the insect (35). When winter types are spring planted, the decumbent growth habit may enhance the chance of escape from beetle attack (48). Adult beetles confined to two leaves of the same plant laid significantly more eggs on the younger leaf. Adults did not dis- tinguish between leaf ages when feeding on 'Clintland 64' oats. Less feeding occurred on the younger leaf of vernalized 'Genessee' wheat and in conjunction with this late planted fall grains and early planted spring grains appear to have more cereal leaf beetle damage (29). Variation in the phenotypes of host plants makes defining resistance difficult. Seedling laboratory tests are used to minimize these differences (35). Twice as many eggs were oviposted on oat seedlings 4.3 inches high as on seedlings at 7.2 inches (55). The angle of illumination and leaf width affect positioning of the eggs on the leaf blade. When the plants were inverted, the number of mis- placed and fallen eggs increased (52). Field collected females exhibited greater egg production efficiency after feeding on oats as compared to wheat (49). Ovi- position preference, reduction in egg production, larval weight gain and survival were observed for the insect when fed on wheat, oat and barley lines (31). Larval weight gain and survival were measured for lines susceptible and resistant to larval feeding damage in field trials. Larval tests substantiated field test results (34). Seedling mortality, after exposure to pre-diapause adult feeding, was highly correlated with larval feeding in the field (47). Adults exhibited differences in oviposition preference when four wheat lines, with different pubescence densities, were compared. Differences in percent egg hatch and larval weight gain were seen. Ten percent of the first instar larvae survived after confinement to densely pubescent wheat for three days (33). Six lines dif- fering in pubescence density and length were studied to observe in- sect reaction. Lines with many long trichomes (30 per mmz, 0.25 mm long) had the fewest eggs per plant and the lowest percent larval survival (19). Oat leaf hairs tend to be longer and less rigid than those of resistant wheat (46). Eggs laid on densely pubescent leaves are more susceptible to desiccation and have low hatch percent. Slow larval weight gain was the result of ingestion of trichomes as the larvae tried to reach the leaf surface (32). First instar larvae bit each trichome as many as three times before reaching the nutritious leaf epidermis. Fourth instar larvae ingest the trichomes whole. Undigested trichomes may penetrate the mid-gut epithelium. Larvae fed on resistant C.I. 8519 grew slowly (51), as did larvae fed P.I. 311677, an A. sterilis line (39). Johnson (20) has examined the environmental effects and ecological importance of plant pubescence. A general overview of pubescence of many species was made. Uniform leaf expansion was important when determining trichome density of small grains (45). The first leaf of seedlings grown in closed petri dishes was more pubescent than those grown in peat pellets (41). Plants grown at high light intensity exhibited fewer trichomes than those grown under low light conditions when temperature and relative humidity were held constant (12). Profuse increases and decreases in trichome density were observed in several Triticum and Hordeum species after chronic gamma irradiation. The Optimum dose for maximum hair production varied with species and variety. Extra visible incandescent light affected hair length and density in_fl. spontaneum. Increased hair- iness was not transmitted to the offspring, but rather appears to be an immediate physiological response of the leaf epidermis (21). The inheritance of small grain leaf pubescence has been studied. In 1968, Ringlund and Everson (28) found that pubescence in wheat is quantitatively inherited and is due to partial dominance and additive effects. Wallace et a1. (43) found 3:1 pubescent to glabrous segregation in F2 wheat populations. Leisle, in 1974, working with durum wheat, described trichome expression as the re- sult of two or three dominant genes for density which are additive for length. All glabrous F2 plants had no pubescent segregates in the F3 (23). Segregates of wild Hordeum sp. indicate that pubescence is dominant but complex in inheritance (13). Resistance to larval feeding damage of barley appeared to be recessive and due to genetic interaction (17). Pubescence in oats appeared dominant or partially dominant, as segregates were predominantly pubescent (25). Sarkarung (30) found two genes for pubescence expression in oats. Modifying genes increased trichome counts of progeny of his crosses over density of the A, sterilis parental lines. Since oat and barley lines have few trichomes, alternative methods for securing resistance may be needed. Oviposition pre- ference was significantly different between substitution lines of Hope wheat (40). Identification of a resistant genome may be useful in oats. Larvae feed between the veins. Feeding scars are parallel to the veins (50). Antibiosis occurs in lines which have veins too close together for larvae to feed (37, 38). Narrow leaves or curly leaf blades may be less preferred for oviposition; leaf cuticle structure and leaf color may also be of importance (46). PLANT MATERIAL The plant materials used in the following tests were ad- vanced generation progeny of a diallele constructed by Dr. Robert Steidl. Dr. Fred Collins of the University of Arkansas contributed eleven pubescent segregates, three of which were used for various crosses. ' The parents for Steidl's diallele were selected on the basis of field screening tests for resistance to the cereal leaf beetle (Smith and Webster,unpublished data). An Avena sterilis L. line, P.I. 311677, was selected for its resistance and leaf pubescence. Of the progeny of the 7 x 7 diallele only those crosses with P.I. 311677 as a parent exhibited any pubescence. These crosses, 5, 11, 12, 13, 14 and 15, were screened in later tests. Lines 5 and 15 were eliminated, Table 1. Nine segregates of Nora X TAM 0-301 from the University of Arkansas were screened for pubescence density. Segregate 105-2, labeled line A-6, was retained for further testing. Two segregates of Nora X NCCR-3 were available and both were retained for further testing. Two crosses in addition to the diallele were made: (Nora X NCCR-3)X P.I. 311677 and (Nora X TAM 0-301)X P.I. 311677. Because the pubescence was from two different A, sterilis sources, genetic complementation could result. 10 TABLE 1. Line Identification Numbers and Pedigree Line Pedigree Pubescence Source 5 'Korwood' C.I. 9167, X P.I. 311677 P.I. 311677 11 'Froker' X P.I. 311677 P.I. 311677 12 C.I. 521 X P.I. 311677 P.I. 311677 13 C.I. 1625 X P.I. 311677 P.I. 311677 14 C.I. 4867 X P.I. 311677 P.I. 311677 15 C.I. 4893 X P.I. 311677 P.I. 311677 A-6 Nora X TAM 0-301 (105-2) P.I. 295919 A-lO Nora X NCRR-3 (114-1) P.I. 320793 A-ll Nora X NCCR-3 (114-5) P.I. 320793 29 A-6 X P.I. 311677 P.I. 295919, P.I. 311677 0 A-lO X P.I. 311677 P.I. 320793, P.I. 311677 E P.I. 311677 P.I. 311677 MAR 'Mariner' C.I. 9195 NONE Crosses for Recurrent Selection Line 11 12 l3 l4 A-6 A-lO A-ll 29 Frequency 6 14 10 NON TESTS Painter (27) described insect resistance in plants as the heritable qualities that influence the host plant-insect inter- action. Plants could be morphologically alike but physiologically different. The tests used to observe cereal leaf beetle resistance in oats were based on three aspects of resistance; preference, anti- biosis and tolerance. Beck (1) described host plant-insect interaction as having two aspects, host selection by the insect and plant resistance to the insect. Resistance components were defined as preference and antibiosis. Beck stated that tolerance was solely a plant char- acteristic not related directly to plant-insect interaction. Lee (22) examined cereal leaf beetle preference in barley lines by oviposition preference tests, antibiosis by larval weight gain and feeding damage, and tolerance by plant recovery. He con- cluded that these aspects of resistance operate independently of each other and inheritance of each is complex. Preference or non-preference is due to plant characteristics which cause insect movement away from the plant for food, oviposi- tion or shelter. Leaf pubescence decreased feeding and oviposition on wheat lines. Without the "right kind" of contact stimulus, oviposition would not occur. The effect of trichomes on adults may be visual or tactile. 11. 12 Preference was evaluated by oviposition preference tests under both field and greenhouse conditions. The test plants were infested with adult cereal leaf beetles. After three days of feeding and oviposition, the number of eggs on each plant were counted. Plants with fewer eggs were less preferred as oviposi- tion sites. The number of eggs and larvae were counted for five stems of each test row in the field. Outdoor tests permitted evaluation of plant material under more natural conditions. Antibiosis occurs when resistant plant material is fed to insects and a decline in viability results. Individual plants vary in suitability for insect food. The nutritional quality of the food is important. Ingestion of high fiber, low nutrient, leaf trichomes resulted in prolonged larval development and slow weight gain. Larvae appear restless when confined to highly pubescent wheat plants. Antibiosis was estimated through larval feeding damage scores. Plants more severely damaged were more suitable food sources. Resistant plants were less acceptable food sources, less damage re- sulted. Tolerance measures the plants' ability to recover after in- sect attack. Replacement and regrowth of tillers is necessary with an infestation of chewing insects like the cereal leaf beetle be- cause damaged plant tissue is beyond repair. Cereal leaf beetle tolerance was not evaluated due to difficulty of determining re- covery. 13 l. Trichome Counts High trichome density is correlated with CLB resistance in 2 reduce adult wheat. Trichomes of .20 - .30 mm in length and 30 mm and larval feeding damage. This association also occurs in certain wild Hordeum species. Evaluationtrftrichome density in oats is a method of screening for CLB resistance in the laboratory. Materials and Methods Several techniques for evaluating relative trichome density have been developed. The leaf blade rolled over the forefinger and held to the light gives an estimate of density. A section of the blade examined with a dissecting microscope permits a more accurate estimation. Light microscope evaluation is more satisfactory in oats because pubescence density is too scarce to evaluate accurately solely with a dissecting scope. After full expansion of the third leaf had occurred, a sec- tion was cut from the mid portion of the blade. A straight cut, perpendicular to the margin, was made near the center of the leaf. Holding the leaf with the adaxial side up, a second cut was made at a 45° angle from lower right to upper left. Cutting in this manner assured identification of the adaxial surface for proper mounting. The leaf section was placed in a vial containing 50 v : 50 v phenol- chloral hydrate. The vial was labeled and corked. Phenol and chloral hydrate crystals were mixed together in equal proportions by weight. A squeeze bottle with a spout was used to dispense appropriate portions into vials. Vials were corked and placed in a cool dark cabinet to prevent deterioration of the 14 solution. Storage life is limited to approximately two weeks. Solution reuse depends on the amount of discoloration that has taken place. Phenol-chloral hydrate removes chlorophyll from the leaf leaving a skeleton of cell walls. Trichomes, which are an extended cell wall, also remain. Clearing takes two days at 90°F, longer at room temperature. The cleared leaf section was rinsed in 85% lactic acid. The sections were mounted on a labeled slide, adaxial side up, in lactic acid. The slide was observed through a light microscope at lOOX power. A field of view was selected adjacent to the midrib. The area of surface in a field of view was 2.717 mmz. Because the leaf tissue was skeletonized, care was used to assure focus on the upper epidermis rather than the lower. Trichomes on the lower surface are located between the two rows of stomata. Trichomes appear between the stomata and veins on the upper leaf surface. Many trichomes are on the veins on both adaxial and abaxial surfaces. Trichomes in each field were measured with a calibrated ocular micrometer, and classified into three groups by length. Long trichomes were longer than .4 mm, intermediate .15 - .4 mm, and short less than .15 mm. This classification system is very natural since trichomes rarely exceed the limits of these classes. With experience, the observer is able to classify each trichome without measurement. 15 The number of trichomes on the vein and between veins for each length classification was recorded. Five fields were observed on each leaf section. Totals were tabulated and transferred to separate data sheets. Trichome Observations - Fall 1977 During the first week of September in 1977 segregates of lines A-6, A-lO, 29 and E were planted to observe trichome expression of greenhouse grown plants, Table 1. Ten segregates of lines A-6 and E were selected. Individual plants of line 29 were harvested from a caged field test and packaged separately. Seed in fourteen of these packages was selected at random for further observation. Line A-lO was not evaluated in the field prior to the laboratory test as there was not enough seed. Twelve seeds were planted for each segregate in 4 inch pots. Germinated seedlings were transplanted into individual pots and labeled. Six plants of each segregate were selected. Plants were labeled number one through six and leaf samples were collected. Collection vials were labeled with segregate and plant number. Numbered plants of each segregate were individually harvested. Leaves of the remaining plants were not sampled. Seed of these plants was bulked for each segregate at harvest. An analysis of variance was calculated for trichome observa- tions, Table 2. Differences between lines were significant at P 5_0.05. Differences between segregates within lines were highly significant at P 5_0.0l. Because there were unequal number of observations in each line, normal multiple range tests were not 16 TABLE 2. Analysis of Variance of Trichome Counts Per Microscope Field of Greenhouse Seed Increased (Fall of 1977) Lines Tested: A-6, A-lO, 29 and E Source df Mean Square F Test Total 229 Lines 3 269.670 4.195* Segregates (Lines) 36 64.286 3.025** Error 190 21.255 Mean Trichome Number *** E 14.763 A-lO 12.47ab 29 12.02b A-6 9.45c * 936% *‘k P 5 0.01 ***P g 0.05 Duncan's Multiple Range Test 17 appropriate. An individual S; was calculated for each pair of - means compared, Table 2. The mean trichome count of lines E and A-lO were not significantly different nor were lines A-10 and 29. Line A-6 had significantly less hairs than the other lines tested. 2. Resistance as Measured by Ovipositional Preference Oviposition tests determine cereal leaf beetle resistance as measured by adult preference. In these tests the adults are free to feed at random and the females lay eggs where they feed. A more favored food source will have more feeding damage and thus more eggs. This test evaluated relative preference within the cages in the green- house or plant growth chamber. Plant size; i.e., height, number of leaves, and total leaf area affect adult contact with plants and relative amount of leaf area available for feeding. The number of eggs oviposited per row sharply influences potential larval feeding damage (LFD) which is dependent upon the number of eggs hatched and the suitability of the host for larval feeding. Materials and Methods Planting took place after the number of lines to be tested were determined. Two holes were punched in the bottom of 5 ounce paper Dixie cups. The cups were filled with steamed soil and the soil tamped. One seed was planted per cup after the cups were labeled with three inch plastic stakes. Three cups planted for each of two lines to be tested is usually adequate. More cups are needed 18 for paired comparisons as in parent/progeny tests to assure an adequate supply of plants. Plants with Avena sterilis germplasm re- quire more pots to accommodate dormancy factors which delay germina- tion. The cups were placed on cafeteria trays (35 per tray). Trays were placed in a warm greenhouse set at 80°F with a 16h photophase to optimize germination. As the seedlings emerged, in- dividual Dixie cups were removed to a cool greenhouse or growth chamber set at 60°F with a 16h day and 55°F with an 8h night. The cool temperature slows seedling growth in an attempt to keep rapid and slow germinators near the same size. Ten days after planting a daily tabulation of the number of seeds germinated was made. Seedlings available to test were randomized with the aid of random number tables. Beetle-plant interaction varies greatly. "Hot spots" occur where much feeding and oviposition take place. For this reason, a randomized block design is preferred over completely randomized de- sign. A segregate of each line was randomly assigned to each replication. Two trays end to end fit across the width of the cage with six trays along the length. Four hundred twenty plants can be tested at once in each cage. A susceptible check border to spread beetles was used around the perimeter, limiting the number of test plants to 336 for each test. The twenty replications used were adequate to distinguish differences between lines. 19 The cage dimensions were 7' by 3' by l'. The long sides each had three doors for moving plant material into the cage. The ends and top were made of lumite screen, and the base was plywood. The screen mesh (32 x 32) was small enough to contain adults but per- mit water and light entry. Plants were watered with a mist nozzle through the side doors. After plants were arranged in randomized groups and watered, adult cereal leaf beetles were counted and introduced. One beetle for every two plants produces consistent test results. It was assumed, from test to test, that the male:female ratio was about one to one. This assumption eliminated the need for sex determination which is tedious and time consuming. Laboratory reared beetles require fourteen days of prepara- tion prior to use in oviposition preference tests. Post aestivation adults are removed from cold storage and placed in a small cage with 'Larker' barley seedlings, where they feed and oviposit. Initiation of oviposition occurs after ten days. Maximum egg output is desired for oviposition preference tests. Adults collected from the field were used when available, usually for tests in May. Field collected adults were held in cages in a cold dark growth chamber until test material was ready. Warming the adults and having food available resulted in less feeding damage to test material after introduction into the oviposition cage. Tests during winter months were conducted in the greenhouse at 80°F with a 16h photophase. Cold outdoor ambient temperature permitted cooling of the greenhouse if sunny weather prevailed while 20 testing. Spring tests were conducted in a growth chamber at 80°F, 16h light; 70°F, 8h dark. Adequate cooling in the greenhouse was not available to maintain 80°F through spring and summer months. Adult feeding damage and egg production were observed closely. Temperature reduction slows feeding and prolongs oviposition. Beetle source, field collected or laboratory reared, and previous food available affect oviposition and feeding rate. When feeding damage and eggs were observed on most of the plants the adults were removed and the total number of eggs per plant and the number of leaves per plant were recorded. Plants should be moved to a cool greenhouse or growth chamber 60°F, 16h light, 55°F, 8h dark to re- cover. Cool temperature slows the internal development of the eggs and eventually the eggs turn brown and die. When cool recuperation space was not available, plants were fumigated when larvae emerged. Each pot was labeled with an additional stake denoting ovi- position test number and indicating the replication and position in the cage. After the test was completed the plants were transplanted into 4 inch clay pots by peeling the Dixie cup from the root ball and filling in the pot with additional steamed soil. When plants reached the five leaf stage, the third leaf was cut and cleared in 50v:50v phenol-chloral hydrate as indicated in the trichome section. Statistical analysis followed with calculation of separate analysis of variance for trichome number and egg counts. A correla- tion coefficient was obtained between trichome density and egg count. 21 Special techniques were used when testing parents and progeny of CT'OSSES. Oviposition Test - Spring 1978 On 7 March 1978, twelve Dixie cups of each segregate of lines A-6, A-lO, 29 and E were planted for an oviposition preference test. 'Mariner' was included as a susceptible check. All seed was de- hulled by hand prior to planting. Poor germination of line E was due to dormancy factors typical of A, sterilis. Line E was eliminated from further participation in this test. Because of erratic and late germination of many of the segregates, only part of the plants were tested for oviposition preference. All plants were sampled for trichome evaluation. One hundred eighty-two plants were arranged in a completely randomized design. The plants were placed in the cage in a warm greenhouse. Line 29 was represented by 51 plants, A—6 by 68 plants, and A-lO by 18 plants. Forty-five Mariner plants were included. The position of each segregate was recorded. On 21 March 1978, one hundred post-aestival adult cereal leaf beetles were introduced into the cage. Feeding, mating and oviposition continued for three days at which time the plants were removed and adults collected. The number of eggs on each seedling was recorded with segregate number and cage position. Seedlings were transplanted into 4 inch clay pots and placed in a cool growth chamber to recover. After full expansion, the third leaf was cut and cleared in phenol-chloral hydrate. Leaf samples were evaluated for trichome length and density. 22 An analysis of variance was calculated for trichome counts of plants included in the oviposition test. An additional analysis, including late germinating plants not included in the test, was calculated. Differences in trichome counts between lines included in the oviposition test were significant at P_g 0.01. Segregates were significant at P 5 0.01, Table 3. An analysis of variance was calculated for egg counts. Variation between lines and between segregates were statistically highly significant at P §_0.0l, Table 3. A modified Duncan's Multiple Range was used to test dif- ferences in mean egg count of each line. Mariner had 6.556 eggs per plant. Lines 29, A-6, and A-lO were not significantly different at P §_0.05. Line means were 2.647, 3.013, and 3.0 respectively. The egg count was regressed on trichome count for each line, Table 4. Mariner was not included in regression or correla- tion calculations. A consistent zero for trichome density would upset the relationship between egg count and trichome density for those lines exhibiting variation. The correlation coefficients in Table 4 are an estimate of zero, which indicates a lack of relationship between trichome number and egg laying. This may indicate no cause and effect or it may be due to sampling. In view of this strong relationship in wheat and some information on oats which will be introduced later, I believe that sampling numbers are at fault, or it may be that trichomes may be too few to be biologically significant. 23 TABLE 3. Analysis of Variance of Oviposition Test Egg and Trichome Counts (Spring of 1978) Lines Tested: A-6, A-lO, 29 and Mariner Egg Counts per Plant Source df Mean Square F Test Total 181 Lines 3 154.290 6.707** Segregates (Lines) 27 23.006 4.046** Error 151 5.686 Trichome Counts, Included in Oviposition Test Source df Mean Square F Test Total 168 Lines 3 3848.267 35.873** Segregates (Lines) 26 107.256 1.825** Error 139 58.787 Trichome Counts, Five Microscope Fields Per Leaf Source df Mean Square F Test Total 280 Lines 3 3771.093 59.688** Segregates (Lines) 27 175.330 2.775** Error 250 63.180 ** P.§ 0.01. 24 TABLE 4. Correlation and Regression of Egg Counts on Trichome Number of Oviposition Test (Spring 1978) Mean Egg Mean Trichomes Count Per Line df Regression Correlation Per Field* Plant 29 45 -0.049 -0.l998 3.287b 2.647b A-6 58 -0.028 -0.0760 2.293b 3.015b A-lO 16 0.017 0.1138 4.033a 3.000b Mariner --- --- 0.0C 5.556° *Duncan's Multiple Range Test P_g 0.05. 25 The mean trichome density per field was calculated. Dif- ferences between lines were tested with a modified Duncan's Multiple Range test as the number of observations per mean were not equal. Mariner had no trichomes. A-lO had 4.033 per field. Lines 29 and A-6 were not significantly different, with 3.287 and 2.923 respec- tively. Segregates within lines were found significant at Pu: 0.01, as were lines in analysis over all of the plants, Table 4. Oviposition Line Test - Winter 1979 During the third week of January in 1979 seed was planted for two oviposition preference tests. One test was an overall comparison between lines that had been used in crosses. The other was a direct comparison between parents and their F]. When supplies of seedlings for crosses tests were inadequate, excess plants from the line test were used. Further discussion of the crosses test follows in the crosses section. On January 16 and 17, 1979 lines 11, 12, 13, 14, A-6, A-lO, A-ll, 29, Q and E were planted for the line test. The number of seedlings from each line was tabulated. Plants were arranged in a randomized block design with a representative plant of each line in each of twenty replications. Mariner was included in each replica- tion as a susceptible check. A single border row of cups of Mariner encircled the entire test. A segregate from each line was assigned to each replication at random. On 30 January 1979, plants were placed in position on cafeteria trays and set inside the oviposition cage. The test was conducted in the greenhouse with two hundred twenty test plants and 26 sixty-six border row plants in the cage. One hundred fifty labora- tory reared post aestival adult beetles were introduced. Two days later, on 1 February 1979, after the adults were collected, the plants were removed from the cage. The total number of eggs and plant height in cm was recorded along with line and position for each entry. Individual stakes were made for each Dixie cup indicating replication number and position within the replica- tions. After transplanting, the third leaf was sampled. Trichome length and density were evaluated. An estimation of variance was calculated partitioning the total variance into replication, between line, segregate within lines, and plant variability for both egg count and trichome density data, Table 5. Differences between egg counts of segregates within each line were found significant at P.5 0.01. Differences between replications were not significant. The total trichome counts of five microscopic fields per leaf were tested for significance. Differences between replications were not significant. Differences between lines and between segregates within each line were highly significant, Pig 0.01, Table 5. The correlation coefficient between trichome number and egg count for each line was calculated. Correlation Coefficients ranged from -.63 for line 11 to .39 for line E. Mariner was not included in the correlations as it has no trichomes and would skew the results. Lack of variation in one variable brings the correlation to zero, Table 6. 27 TABLE 5. Analysis of Variance of Oviposition Line Test (Winter of 1979) Lines Tested: 11, 12, l3, l4, A-6, A-lO, A-ll, 29, Q, E and Mariner Egg Counts per Plant Source df Total 219 Replications 19 Lines 10 Segregates (Lines) 32 Error 158 Mean Square F Test 15.363 0.811"5 87.383 4.612** 18.946 1.488* 12.729 Trichome Counts, Five Microscope Fields per Leaf Source df Total 219 Replications 19 Lines 10 Segregates (Lines) 32 *9.5 0.05. *1: P.: 0.01. Mean Square F Test 1353.946 1.042"5 9106.721 7.006** 1299.836 19.047** 28 TABLE 6. Correlation and Regression of Egg Count on Trichome Number from Oviposition Preference Test (Winter of 1979) Mean Trichome * Mean Egg Number*per Line Regression Correlation Count* Field 11 -O.16l -0.63l 4.45Cd 4.74de 12 0.041 0.155 5.45de 2.69e 13 0.011 0.029 8.70a 5.54d 14 -0.011 -0.056 6.25ade 2.24ef A-6 0.082 0.326 6.65abc 6.41Cd A-lO 0.002 0.030 3.95d 11.48a A-ll -0.086 -O.546 7.05ab 2.69e 29 -0.015 -0.083 7.25ab 9.05b 0 -0.089 -0.273 7.65ab 8.17bc E 0.044 0.390 1.656 12.25a Mariner 0.0 0.0 4.10Cd 0.00f * 18 df. *9: Duncan's Multiple Range Test P‘: 0.05. 29 Egg counts were regressed on trichome density. Regression Coefficients ranged from .028 to .358. Regression Coefficients are presented in Table 6. A Duncan's Multiple Range test was used to test significance of line differences of mean egg count and mean trichome count per field. Line E had the fewest eggs and the most trichomes. Line 13 had the highest egg count mean of 8.7. Line 14 had the fewest trichomes with 2.24 per field, Table 6. 3. Larval Feeding Damage Larval feeding damage observations show the reaction of the cereal leaf beetles to plant material in the field under natural conditions. The intensity of larval feeding damage (LFD) is a function of the interaction between the host and the insect. Three cages were used to establish three levels of infestation. Materials and Methods Large LFD cages (10' x 9' x 5') were constructed in the field each year. Each cage was designed to accommodate 600 rows. Six rows ran across the width, planted end to end, and were each 1.5' long. One hundred rows, 9 inches apart, were planted the length of the cage. The cage size may be increased if needed (Smith and Webster, unpublished). After soil preparation, a rolling barrel marker was used to mark the rows (6). Coin envelopes of seed were hand planted. The cage was constructed after the seedlings emerged. 30 Nine 10' two-by-six inch boards were laid on edge in shallow trenches and staked. Soil was filled in along the boards to pre- vent beetle escape. The boards forming the length of each side were held together end to end with metal brackets. Another 10' board was nailed across the cage at each end. Two curved pieces of aluminum conduit were fitted together to form an arch. An arch was anchored to each bracket where two boards joined. Three pieces of wire were threaded through hose clamps and placed on the arches to string them together. The wire was stretched and the hose clamps tightened together to form a secure framework. The wire was secured to the boards at each end. Two 9' x 100' tobacco plant bed covers (Grade 3A) were sewn together to form an 18' x 100' sheet. The cover was stretched across the cage framework and stapled to the boards along the sides and across the southwest end. Excess material at the north end was rolled up on a two-by-four to close the cage but still permitted re- opening for entrance into the cage. Post-aestival adult cereal leaf beetles collected in the field (7) were counted in the laboratory, prior to release in several locations inside the cages. The adults were free to fly around and select plants at ran- dom for feeding. Upon landing on non-preferred plant material, the beetles resumed flying in search of a more favorable food source. Feeding occurred intermittently with oviposition. The female ovi- posits where she feeds. Susceptible plants received more eggs. 31 The attachment of eggs was poor. Eggs haphazardly perched on the tips of trichomes frequently desiccate or are blown or washed off by rain prior to hatching. The cages were removed when larval and egg counts reached the maximum. The plant bed cover shaded the plants and was removed at the earliest possible date. Egg and larval counts were made at this time. Five tillers were selected at random within a row. The number of eggs and number of larvae were recorded. Analysis of variance was calculated to test significance of differences between lines. Five days after oviposition the larvae emerged. Emergence was delayed or inhibited by cool temperature. The larvae began feeding on the leaf where the eggs were laid. Movement from leaf to leaf in search of food occurred. The larvae can move from plant to plant. Increased movement of larvae and less feeding resulted on resistant plant material. After an extended period of feeding, larval feeding damage was evaluated and recorded. A scale of one to nine was used. One indicates little feeding damage, less than 10% of the leaf consumed. Two indicates 20% of the leaf consumed, etc. A distinction must be made between adult and larval feeding and damage. Adult feeding penetrates the leaf causing shredding. Larvae strip the upper surface, usually leaving translucent skeletal tissue intact. Additional data such as height, maturity and growth habit may be evaluated to determine any effect on larval feeding damage. 32 An analysis of variance was calculated for each trait. Multiple regression may be used to determine the relative importance of each trait measured and influence on larval feeding damage. Larval Feeding Damage Field Test - Summer 1977 In May 1977 Dr. Robert Steidl, a post-doctorate with the project, planned and planted a larval feeding damage test at the MSU Crops Farm. The data and resulting analysis are from my own observa- tions of this material. Thirty lines were planted in twenty replications. Replica- tions were arranged to minimize variation due to soil and light gradient differences. The lines planted were segregates and parents of Dr. Steidl's diallele. Lines A-6 and 29 were also included for field evaluation. Lines A-6 and 29 were not included in diallele crosses. The cage was constructed when the plants had approximately three leaves per tiller. On 23 May, 3,500 field collected adults were introduced into the cage. Five hundred additional beetles were added on 27 May, to replace those which had escaped or died. The cage was removed on 4 June. Thus, the beetles were confined to the test plants for twelve days. Five plants from each row were selected at random. The third leaf was cut and cleared in 50v:50v phenol-chloral hydrate. Only one sample per plant was made for rows with less than five plants. Vials were labeled for replication and row number rather than line number to prevent bias when trichome evaluations were made. 33 The trichomes were measured and classified by length. An analysis of variance was calculated. Replications and lines exhibit significance at P's 0.01, Table 7. An analysis of variance was calculated using only pubescent lines. The F Tests exhibit significance at Pig 0.01 for both replications and lines. Because of this significance, replications and replication data will be used for segregates of each line in future tests. Trichome number per field was averaged over twenty replica- tions. Means of pubescent lines were compared using Duncan's Multiple Range. The analysis of variance and results of the range test appear in Table 7. Lines with the same letter are not significantly different at P_g 0.05. Larval feeding damage was evaluated on July 1. The scale of 1 through 9 described previously was used. An analysis of variance over all of the lines was calculated. Replications and lines ex- hibited significance at P‘s 0.05, Table 7. An analysis of variance for trichomes using only pubescent lines also found highly signifi- cant differences between lines and between replications, Table 8. A Duncan's Multiple Range test of line mean larval feeding damage scores over twenty replications found lines not significantly different from each other at P‘s 0.01. Avena sterilis line E, stood alone as the most resistant entry in this test, Table 9. Plant growth habit was also evaluated on July 1. Decumbent plants may miss beetle attack. A scale of one to five was used to classify decumbency with one as decumbent and five as upright. All of the A, sterilis plants were scored as one. The remaining parents 34 TABLE 7. Analysis of Variance Using All Thirty Lines From Field Data (Summer of 1977) Trichome Frequency per Microscope Field Source df Mean Square F Test Total 599 Replications 19 0.7197 1.966** Lines 29 15.4697 42.267** Error 551 0.3660 Larval Feeding Damage Source df Mean Square F Test Total 599 Replications 19 5.4718 6.116** Lines 29 3.8982 4.357** Error 551 0.8947 Decumbency Source df Mean Square F Test Total 599 Replications 19 0.3137 1.407”5 Lines 29 13.1942 62.533** Error 551 0.2229 *1: P §_0.01. 35 TABLE 8. Analysis of Variance of Pubescent Lines From Field Test (Summer of 1977) Trichome Counts, Five Microscopg_Fields_per Leaf Source df Mean Square F Test Total 179 Replicates 19 2.333 2.076** Lines 8 13.154 11.706** Error 152 1.240 Larval FeedingADamage Source df Mean Square F Test Total 179 Replicates 19 4.576 4.659** Lines 8 4.476 4.557** Error 152 0.982 Decumbency Source df Mean Square F Test Total 179 Replicates 19 0.3096 0.427”5 Lines 8 30.800 42.457** Error 152 0.725 ** P‘: 0.01. 36 TABLE 9. Duncan's Multiple Range Test of Pubescent Line Means From Field Test (Summer of 1977) Mean Trichome* Larval* * Lifl§_ Number/Field Feeding Damage Decumbengy 5 .323f 4.6” 4.5” 11 1.489C’d’e 4.15” 4.45”’” 12 1.38d’e 4.6” 4.85a 13 1.377”:9 4.1” 4.85a 14 2.168b’d 4.0” 3.45c 15 .9719:f 4,2b 4.3a,b E 1.755”’”’d 3.0a 1 0d 29 3.142a 4.16” 0c A-6 2.253” 4.35” 4.2” *Plg 0.05, Duncan's Multiple Range. 37 and many of the progeny exhibited upright growth habit. A, sterilis hybrid progeny exhibited segregation for growth habit. An analysis of variance of all thirty lines was calculated. An F test found variation between lines highly significant at P 5_0.01, Table 7. When growth habit of the pubescent lines was analyzed separately, replication mean squares were positive but not significant at P 5_0.01. A highly significant difference between pubescent lines was observed, P §_0.01, Table 8. The results of a Duncan's Multiple Range test are also shown in Table 9. Larval feeding damage was regressed on trichome count and growth habit. Growth habit was correlated with trichome count, Table 10. Indivi- dual replication observations were used for the basic unit. Each correlation coefficient was based on twenty paired observations. Each line was harvested by rows, except for line 29, which was har- vested by individual plants. Three Levels of Infestation - Summer 1978 During the second week of May, seed was planted in the field for three cages which would be used to simulate three levels of beetle infestation. Ten segregates of six lines (11, 12, l3, l4, A-6, and E) plus Mariner, a susceptible check, were included in each of ten replications. In addition, line A-ll was included in only six of the ten replications due to limited seed. Lines were randomized within replications independently. Segregates are confounded with replication because the segregates occur in the same replication in each cage. 38 TABLE 10. Correlation and Regression Between Traits Measured During the Field Test (Summer of 1977) Regression Regression Correlation of of LFD on * of LFD on * Trichome Count * Line Trichome Count Growth Habit and Growth Habit ll -O.106 0.028 0.028 12 ~0.04l —O.264 0.036 13 0.412 -O.374 0.388 14 0.169 0.183 0.523 E -0.045 0.0 0.0 29 -0.141 -0.022 -0.039 A-6 0.628 -O.269 0.041 * 18 df. Trichome Count = Mean of Five Microscope Fields per Leaf. 39 On 31 May, the three separate cages were constructed. One thousand, 2,000, and 4,000 field collected adult cereal leaf beetles were introduced to simulate a density of 1, 2, and 4 beetles per square foot of cage floor area respectively. On 22 June, when eggs and larvae were visible on the plants, the low level of infestation cage was removed. The intermediate and high level cages were removed the following week. Larval feeding damage was evaluated on 4 July. A scale of 1 through 9 was used. An analysis of variance partitioned total variation into variation between levels of infestation, replica- tion by infestation interaction, differences between lines, and line by infestation interaction. The residual variation was used as the error term. Levels of infestation were found to be statistically significant at P 5 0.01. Variation between lines and the line by infestation level interaction term are highly significant at P 5_0.Ol, Table 11. A Duncan's Multiple Range test was made to test significance of differences between lines. Line E experienced the least larval feeding damage; Mariner the most. The results are presented in Table 11. Three Levels of Infestation - Sumner 1979 -0n 7 May, seed was planted for simulation of three levels of infestation. Ten replications of eleven lines were planted in each cage. The lines included were 11, 12, 13, 14, A-6, A-lO, 29, Q, and E. Mariner was also included as a susceptible check. On 23 May, the three cages were constructed over the test plants and on 25 May field collected adult cereal leaf beetles were TABLE 11. Analysis of Variance and Comparison of Larval 40 Feeding Damage Scores (Summer of 1978) Source df Mean Square Total 207 Infestation 2 41.94 Rep (Infest) 27 6.34 Line 6 14.46 Line (Infest) 12 9.12 Error 162 2.43 Line Mean* E 5.86a A-ll 6.19”” 13 6.56abc A-6 6.98”Cd 11 7.06”Cd 14 7.17Cd 12 7.56Cd Mariner 8.03d *P 5_0.05, Duncan's ** P g_0.01. Multiple Range Test. F Test 6.32** 5.96** 4.00** 41 introduced into the three cages. The cage covers were removed on 20 June. The number of eggs and larvae were recorded for 5 tillers/ row on 20 and 21 June. Larval feeding.damage was evaluated on 28 June using a 1-9 scale. Plant height and growth habit were also measured on 28 June. Height measurement of plants in cage 2 was delayed until 2 July. Maturity observations were made on 6 July. A scale ofl -5 was used. One represented plants in the seedling stage, 2; tillering, 3; heads in the boot, 4; heads extruding, and 5 when heads had fully emerged. A correlation matrix was calculated for all of the traits observed. When larval feeding damage was correlated with plant growth habit, height, egg count, larval count, egg plus larval count, and maturity, the correlation coefficient was near .5 in each case. Plant height, growth habit and maturity are highly correlated with each other as are egg count, larval count and egg plus larval counts. Decumbency was associated with plant height as plant height was associated with maturity. Egg count plus larval count or "Count"* was a direct function of egg counts and larval counts. "Count" is more highly correlated with larval feeding damage, height and growth habit than either larval count or egg count alone. An analysis of variance was calculated to determine the level of significance of variation between lines tested and between levels * "Count" equals egg count plus larval count. 42 of beetle infestation. Differences in egg counts, larval counts and "Count" were highly significant, P 5_O.01, between both lines and cages. Differences between lines were significant for both maturity and growth habit although variation between cages was not signifi- cant. Unifbrmity between cages and variation between lines was as expected. Variation in height was significant between lines and levels of infestation. Differences between lines were expected. The plants in cage 2, the intermediate level of infestation, were significantly taller. Height measurements in cage 2 were made four days after cages 1 and 3 due to adverse weather conditions. The mean "Count" of cages 2 and 3 was not significantly dif- ferent. The plants in cage 3 had more eggs and larvae but not significantly more; subsequently, more larval feeding damage re- sulted. Larval feeding damage to the plants was significantly dif- ferent between the three cages. Several multiple regression equations were calculated to explain the relative importance of each trait observed for estima- tion of larval feeding damage. Larval feeding damage was regressed on lines and growth habit. Growth habit explained more variation in larval feeding damage than did lines. Variation between lines was not significant at P 5_0.05 by F Test. Growth habit was highly significant, P < 0.01. Larval feeding damage scores were regressed on levels of infestation, "Count" and line. Cage and "Count" were highly 43 significant P < 0.01. Lines were not significant at P 5_0.05 by F Test. Larval feeding damage scores were then regressed on all of the variables. Level of infestation, egg count, larval count, and plant height were highly significant in estimating larval feeding damage.” Growth habit was significant at P 5_0.Ol. Lines and maturity were not significant. Lines were significantly different for larval feeding damage. The analysis of variance for larval feeding damage scores and "Count" appear in Table 12. Line means were compared by Duncan's Multiple Range Test, Table 13. 4. Recurrent Selection Experiments The purpose of within line crosses was to allow re-assess- ment of genes for trichome expression. Entries were selected from the World Oat Collection and crossed. Progeny of these original crosses were selected, those retained were then tested for various aspects of cereal leaf beetle resistance. Pubescent segregates of each original cross were intercrossed within each line to bring to- gether again genes from the original parents. When segregates possessing different alleles for trichome expression are inter- crossed,. an increase in trichome length and density should occur. Materials and Methods Trichome density and field larval feeding damage scores were determined during the summer of 1977. Advanced generation segregates of progeny of Steidl's diallele set of crosses 44 TABLE 12. Analysis of Variance of Larval Feeding Damage and "Count" (Summer of 1979) Larval Feeding_Damage Source df Mean Square F Test Total 329 Level Infestation 2 135.545 20.09** Replicate (Level) 27 6.747 Lines 10 28.275 87.03** Lines (Level) 30 10.917 33.60** Error 259 0.325 "Count" Source df Mean Square F Test Total 329 Level Infestation 2 3026.722 26.60** Replicate (Level) 27 113.782 Lines 10 184.193 39.09** Line (Level) 30 225.225 47.79** Error 259 4.713 *3! P 5 0.01. 45 TABLE 13. Comparison of Means for Larval Feeding Damage and "Count" (Summer of 1979) Line Larval Feeding Damage” ”Count"* 11 3.77”” 14.67”e 12 4.97” 16.03fg 13 3.57” 14.07””e 14 3.97”” 15.90fg A-6 3.73”” 14.47”e A-lO v4.03” 15.07ef A-ll 3.80”” 13.40”” 29 3.67”” 11.93” 0 3.72”” 13.15” E 1.63” 7.73” Mariner 5.60e 16.709 *P 5_0.05 Duncan's Multiple Range Test. Mean over Three Levels of Infestation. 46 possessing more trichomes and lower larval feeding damage scores were intercrossed. Within line crosses for several lines werelnade. Trichome counts and larval feeding damage scores were com- pared. Segregates within each line were ranked according to de- creasing trichome counts and separately for low larval feeding damage. 0f the twenty segregates in each line, the three or four with the most trichomes and lowest larval feeding damage scores were selected for intercrossing. When seed quantity was insufficient, the next best segregate in the ranking was selected. Selection of segregates of lines 29 and A-10 was based on trichome data from seed increased during the fall of 1977. Data used for selection is presented in Table 8. Segregates of line A-ll were used without previous test data. Selected parental lines were planted in 4" clay pots during consecutive weeks to increase opportunity for "nicking". Many of the lines were decumbent and tillered profusely. Initially four seeds were planted per pot. As the first crosses were made, it became apparent that fewer plants per pot would be easier to work with. Three seeds were planted in the remaining pots. Occasionally poor germination occurred. This was due to poor seed quality or dormancy from the A, sterilis parent in the cross. After coleoptile emergence, pots were moved to a growth chamber set at 16h light at 60°F with 8h dark at 55°F. When the fifth leaf had emerged from all of the seedlings in the pot, a portion of the third leaf was cut and cleared in phenol-chloral hydrate. The pot was then moved to the greenhouse. 47 Plants of the same line were placed together for identification for crosses. Leaf hairs were measured for length, counted, and tabulated for five microscopic fields per leaf. Significance of trichome counts for line X planting date interaction was tested. Because this interaction term usually serves as the error term an estimate of error was needed. A weighted mean was calculated for treatments. Within treatment sum of squares, the error term Was found by sub— traction. This preliminary error sum of squares, 5:, divided by the simple harmonic mean 110 was needed for an estimate of error and retained error degrees of freedom. The estimate of error mean square was calculated. The unweighted means were calculated for total sum of squares. Variation due to lines and planting dates was calculated. Line by planting date interaction sum of squares was found by sub- traction. The estimate of error from the preliminary analysis of variance was used as the denominator of F tests. Crosses were made by the approach method (2, 5, 8, 24, 26). A five inch wide piece of aluminum foil wrapped on the outside of the dialysis tubing slows desiccation of the emasculated female head. Seed set was poor, as may be expected for greenhouse crosses. Two weeks after pollination, the male parent and hybrid heads were harvested. A tiller selected from the female plant was harvested for seed. Heads were packaged separately and dried. 48 Parents and progeny of crosses were compared by the ovi- position preference test and trichome evaluation. If transgressive segregation had occurred, some of the progeny could be used to test the relationship. A test of the variance of the difference, 53, 0f mid-parent and progeny indicated if a significant change had taken place. Crosses - Spring 1978 Segregates of eight lines (11, 12, l3, l4, A-6, A-ll, 29, and 0) were selected for spring crossses. Four pots per line with four seeds per pot were planted during five consecutive weeks beginning 20 March. The pots were moved to a cool growth chamber after germination in a warm greenhouse. Half of the plants of each line for each planting were transplanted to the field for use as pollen parents. Eighty pots of plants were planted on two foot centers. The four plants of each pot were knocked out of the pot together and planted with the root ball intact. Each clump of plants was labeled with line, segregate number and planting date. After full expansion, the third leaf was cut from all of the plants remaining in the growth chamber. The leaf section was cleared in 50v:50v phenol-chloral hydrate. Three hundred eighty- three plants were sampled. Plants were moved to the greenhouse after the leaf sample was taken. Crosses were made in the greenhouse using the approach method. High temperatures caused dehiscence of pollen while the head was far down in the boot. Opening and clipping of the 49 florets while very young increased the drying. N0 seed set occurred. During the first week in June an outdoor greenhouse was constructed of 2" x 4" lumber and lath snowfencing. Plants in the lath-house grew at ambient temperature and were exposed to varia- tion in weather conditions. Florets were more mature at pollen dehiscence and more pollen was visible on the inside surface of the dialysis tubing than greenhouse grown plants. A modification of emasculation and clipping-techniques was made. Heads with the uppermost floret already dehiscing were selected. The ten most mature florets were used. The remaining florets were cut off. Lemma and palea were clipped shorter exposing more of the stigma. A low percent seed set resulted. Because the outdoor greenhouse was also used for entomological studies, pesticide dust was not permitted. Misinformation on formula concentration resulted in an overdose of dimethoate, a systemic insecticide. The entire parental population died. Transplanted plants were not treated with dimethoate. Good seed set permitted replanting in the greenhouse in the fall. A weighted mean for trichome counts was calculated for variation between the sum of 5 microscopic fields per leaf for forty treatments (eight lines by five planting dates). The pre- liminary error sum of squares was found by subtraction. This estimate of error retains the 343 degrees of freedom for calcula- tion of estimate error mean square. 50 The f0rty unweighted treatment means were used for total sum of squares with 39 degrees of freedom. Trichome variation be- tween the eight lines and five planting dates was calculated. The line by planting date interaction was found by subtraction. Using this interaction term an F test was calculated. Planting date mean squares were not statistically significant but lines were at P_g 0.01. Trichome density variation between lines, between planting dates and the interaction term exhibit significance at P 5_0.0l when an F value is calculated using the estimate of error term for preliminary analysis of variance, Table 14. Crosses - Fall 1978 Seed harvested from plants transplanted for crosses was dried and dehulled by hand. Seed from nine lines (11, 12, 13,14, A-6, A-lO, A-ll, 29, and Q) was planted during seven consecutive weeks beginning on 1 August. Five pots were planted with three seeds per pot for each line. Plants were germinated in the greenhouse and moved to the growth chamber to encourage vegetative growth. The third leaf was sampled from each plant in a pot. The leaf section was cleared and trichome length and density evaluated. When the pots were returned to the greenhouse, cool fall weather had begun. Temperature was controlled by allowing cool air from outdoors to enter the greenhouse. Pollen dehiscence was de- layed until after the head had emerged from the boot. 51 TABLE 14. Analysis of Variance of Trichome Counts per Microscope Field for Eight Lines and Five Planting Dates Lines Tested: 11, 12, 13, 14, A-6, A-ll, 29 and Q Prelimingry Analysis of Variance Source df Mean Square F Test Total 382 Treatment 39 2115.348 13.083** Error 343 161.686 Error Estimate = SZ/no = 554.3843/O.1576 Unwejghted Mean Analysis of Variance Source df Mean Square F Test F Est. Total 39 Lines 7 384.353 3.582** 15.08** Planting Date 4 133.284 1.24nS 5.23** Planting Date x Line 28 107.312 4.21** Error Estimate 343 25.482 ** P §_0.01. 52 Plants of the same line were placed together on the green- house bench. Plants within each line were systematically observed. Those ready for crossing were prepared and the cross set up. Crossing began 2 October. Each line was observed and crosses made twice a week. Approximately the same number of crosses were made within each line. The final cross was completed on 17 December. Parent and Progeny Relations for First Cycle Recurrent Selection- Winter 1979 On 17 and 18 January, parents and a portion of the F1 seeds were planted for each successful cross. Dehulled seed was planted in Dixie cups and placed in a warm greenhouse until germination. The number of Dixie cups planted to parents was proportional to the amount of F1 seed available. Complete crosses, those that had both male and female par- ents and F1 emerged at time of testing, were given top priority for inclusion in an oviposition preference test. Parents were placed on either side of the F]. The three plants were randomly assigned a position together in a row. The cage was then filled with assorted plants that were available. The additional plant material gave the beetles more selection and greater leaf surface area to feed on. The purpose was to test the difference between parents and progeny within each cross not to distinguish between crosses. A total of 196 plants were included in the test. Nineteen were complete crosses, 57 were plants to be used for parent progeny comparison. A border row of Mariner was included. 53 On 31 January 100 laboratory reared adult beetles were in- troduced into the cage. Due to cloudy weather, feeding and ovi- position activity was poor. On 1 February fifty additional beetles were introduced. The beetles were removed on 2 February. Egg number and plant height in cm. were recorded. The third leaf of each plant was sampled for pubescence evaluation. Parents and progeny of crosses were retained for trans- plant and seed collection. The mid-parent, the mean of male and female parents, was calculated. A "t“ test of the variance of the difference between mid-parent and F1 was negative for egg counts. The variance of the difference was not significant by “t" test. The F.‘ plants had fewer eggs than the parental mean. Plants tested for oviposition preference were evaluated for pubescence density. The mean difference of trichome counts when parental mean was subtracted from progeny was positive but not significant by "t" test. The F1 plants exhibited a slight in- crease over the parental mean. Genes for trichome expression are believed to be dominant. Crosses may be successful due to increased gene dosage. The correlation between parent and progeny trichome counts was .911. The correlation coefficient between parent and progeny egg counts was -.055. 5. Arkansas-11 Segregates The oat line Arkansas-ll typifies the effect that leaf pubescence has on host plant-beetle interaction. Arkansas-11 54 segregates All-l and All-3 originated from the same cross. Segregate All-1 is glabrous as are all its progeny. Segregate All-3 and progeny are highly pubescent. Pubescence density is variable. Total trichome counts for five microscopic fields per leaf range from sixteen to ninety-four. While lines isogenic for the gene or genes controlling trichomes are not available, two random isolates of an F3 line, originating from the cross Nora x NCCR-3 (one of which has trichomes and one of which does not) are available. Pheno- typically the lines are alike. Glabrous All-l averaged 8.23 eggs per plant in the oviposi- tion line test conducted during the winter of 1979. The mean egg count of All-3 was 4.86 eggs per plant. When offered seedlings of the same size, the female beetles preferred glabrous plants for ovi- position over pubescent plants. Five replications of All-1 and All-3 were planted in each field cage for larval feeding damage evaluation during the summer of 1979. An analysis of variance separating the two segregates was calculated for each trait observed. Differences between cages was not significant for growth habit and maturity. The plant height in cages 1 and 3 was not significantly different. However, plants in cage 2 were taller because height measurement was made at a later date. Egg count, larval count and egg + larval count or “Count" were not significantly different between All-l and All-3. There were more eggs and larvae on All-1 than All-3 but not significantly more. Adults did not distinguish between the two segregates for oviposition 55 preference in the field. Yet, the larval feeding damage scores on All-l were significantly higher than on All-3. All-3 is a poor food source. The larvae may experience an antibiosis effect as the re- sult of feeding on pubescent All-3. Less larval feeding damage re- sulted. Thus two isogenic segregates had significantly different larval feeding damage scores due solely to differences in pubescence density. Glabrous and pubescent segregates of other lines have not been compared for oviposition preference or larval feeding damage. The results of this comparison justify oat plant selection based on trichome density. Pubescence is biologically significant in the determination of larval feeding damage in oats. 56 DISCUSSION The exact combination of plant characteristics which impart resistance in oats to the cereal leaf beetle is not known at this time. Plant height, growth habit and maturity were evaluated in addition to egg and larval counts and larval feeding damage to ob- serve plant growth rate and size relationship with the beetle. Dif- ferences in plant size and growth habit appear to be more important in the determination of oviposition preference and larval feeding damage than the differences in germplasm. The beetle interacts with the whole plant. Germination percent and rate of germination affect the number of plants and consequently insect contact with that particular line. An abundance of seedlings early on, when plant material is scarce, may render those lines more vulnerable to insect attack. On the other hand, adults prefer young succulent tissue. Newly emerged plants and leaves may be more susceptible. Plant height and number of leaves at the time of infestation may be important. Additional evaluation of seedlings could reveal the influence of plant growth stage on insect attack. If plant height affects ovi- position preference, differences in seedling height may be disguised by differential growth rate at maturity. Differences in oviposition preference for vernalized and non-vernalized wheat seedlings have been observed. The internal physiological change from vegetative to reproductive condition occurs after infestation. Such a change may affect larval feeding rate and damage. 57 All comparisons of larval feeding damage and oviposition preference made in these tests are relative between lines tested. Painter (27) states that preference may exist when susceptible and resistant varieties are grown together. But that resistance may be lacking when the "non-preferred" variety is grown alone. Differences in height and growth habit exhibited between lines are important when beetle reaction is evaluated in a hetero- geneous situation with various lines. Identification of the effects of these morphological differences are important. Selection must be made with regard to, but in Spite of, growth habit, height and maturity. Field crops grown in monoculture are uniform in height and maturity. Beetle preference is then, not a choice between lines, but, whether to feed or not, to oviposit or not, to remain in the oat field or to find alternate hosts for oviposition and feeding. Selection is based on discouraging feeding and oviposi- tion within the field and severe antibiosis to reduce larval growth and feeding damage should adult feeding and oviposition occur. Re- sistance of decumbent or late maturing types based solely on avoidance in the presence of taller types is not effective when plants are grown in monoculture. Methods of testing prospective plant material under mono- culture have been developed by Dr. J.A. Webster. Several lines are tested in individual cages with equal levels of infestation. Cages with a resistant or susceptible check in addition to the test line are used to compare relative beetle activity of lines in separate cages. Selection of resistant lines based on monoculture tests 58 project a more accurate estimate of beetle reaction to the plants under field conditions. Pubescence was evaluated concurrently with other plant characters in an attempt to identify the role of leaf pubescence in oat plant cereal leaf beetle resistance interactions. Isogenic segregates of one line which differ only in pre- sence of leaf pubescence were compared for three levels of beetle infestation. The pubescent segregate had fewer eggs and larvae than the glabrous segregate. Mean larval feeding damage feeding scores were significantly lower for pubescent plants when compared to glabrous plants for each level of infestation. This was the first solid evidence that pubescence affects the cereal leaf beetle reaction to oats. Further exploration will indicate which phases of the insect life-cycle are most affected by dense pubescence. The presence of leaf pubescence has been shown to affect all aspects of adult and larval interaction with the plant material. Many insect species have sensors on the legs, antennae or ovipositor which sense the presence of pubescence upon landing on the plant (27). Leaf blade pubescence is a deterrent to the adults indicat- ing a poor food source and oviposition site. Gravid females con- fined to pubescent wheat plants appear to be nervous (32). Many re- fuse to oviposit on the leaves. Eggs placed on the leaves are hap- hazardly attached. Eggs on pubescent leaves have a high frequency of desiccation and drop off rate (52). The pubescence is distress- ful to the female. Either the female cannot reach the leaf surface 59 or is unable to find enough non-pubescent leaf surface to attach the egg firmly. Oviposition and eclosion take place over an extended period of time in field tests. A single count of eggs and larvae gives no indication of the total number of eggs oviposited nor the drop off rate. Hatch percent and dropoff rate would give more insight into ovipositional preference and plant characters affecting the size of the larval population prior to initiation of larval feeding. Evaluation of hatch percent and dropoff rate would require observa- tions of field plants several times during the oviposition and larval feeding period rather than once in the growing season. Pubescence density differs between plants grown in the greenhouse, growth chamber and in the field. Fedak (l2) found that high light intensity decreased pubescence length and density of barley plants relative to plants grown in low light conditions. Sarkarung (30) found that field and greenhouse growth conditions differed in light intensity and quality, particularly blue, red, far red and ultra violet quantity. Low light may cause etiolation of leaf cells including trichomes. Low light conditions may be cooler, not drying the plants as rapidly permitting more water for cell expansion. High light and consequently higher tempreature may cause smaller, thicker cells. Plants treated with gamma irradiation experienced erratic changes, increase and decrease, in pubescence density. Changes in pubescence density were not inherited suggesting a physiological change in the epidermal cells rather than mutation of the germ cells. 60 Altered growth conditions may be used when growing plants for selection. Extreme light conditions may affect trichome expres- sion of test lines differently. Adverse conditions may be used to distinguish lines with genetic potential for high pubescence den- sity from those with potentially less pubescence. Selection based on low light conditions could distinguish lines that appear identical under "normal" light regime. Selected segregates returned to normal growth conditions may exhibit increased pubescence density due to elimination of excess material that is less pubescent. Distance between vein spacing in leaves may affect suit- ability for larval feeding. Older larvae with wider mouth parts are unable to feed between the veins and have poor weight gain due to ingestion of high fiber leaf vein tissue (54). Webster (44) in- dicated that vein spacing and leaf curl may influence larval feeding damage. Esau (9) observed bulliform cells, "cells shaped like bubbles", between veins of the Gramineae. These highly vacuolated cells are poor in solids content but rarely contain tannins or crystals. Bulliform cells control unrolling of new leaves and hygroscopic opening and closing of mature leaves. By description, larvae feed on bulliform cells and the parenchyma and mesophyll layers below. Selection for smaller, tougher bulliform cells could discourage feeding. Smaller cells could result in slightly curled leaves which Webster (44) proposed to discourage larval feeding. The initial segregate selection was based on both larval feeding damage scores and trichome density. Subsequent selection has emphasized trichome density more so than larval feeding damage. 61 Lack of increased resistance of selected segregates may be attributed to lack of selection for host-plant interaction components. Simultaneous selection for high density and low larval feeding damage scores retains resistance in the next generation. Within line crosses between selected segregates were made. Progeny tests were made using one or two F1 plants and four to ten F2 plants. Due to small population size, the probability of re- combination and reassortment to produce wide differences between progeny was reduced. Differences found between parents and progeny may be due to chance alone. The F1 data indicate that pubescence and ovipositional non-preference are dominant. Segregation in F2 progeny make conclusions difficult but possibly more accurate due to larger population size. Selection of segregates for use in crosses was based on F3 and F4 data of each parental line. Selected segregates exhibited high pubescence counts and low larval feeding damage. Segregates may not differ in genes for trichome expression and oviposition pre- ference. Transgressive segregation occurs with complementation of different genes. The desired increase in trichome number may not be possible through within line crosses. Sarkarung (30) found pubescence was due to gene dosage of one or two partially dominant genes dependent on the line tested. Nora X NCRR-3 segregated 3:l for leaf pubescence with an additional gene for sheath pubescence. Nora X Tam O-30l also segregated 3:l for leaf blade pubescence but contained only recessive genes for a glabrous leaf sheath. Glabrous plants had only glabrous progeny. 62 Highly pubescent F2 plants had only highly pubescent progeny. Lightly pubescent F2 plants segregated into highly pubescent and lightly pubescent plants in the F3. Possibly the Population size was too small to observe glabrous segregates. Modifying genes were found in Tam 0-30l and NCRR-3. These progeny were more pubescent than the A, sterilis from which pubescence was derived. 63 CONCLUSION These tests exhibited that improved resistance to the cereal leaf beetle can be achieved through hybrization and selection. Oviposition preference tests adequately determined adult preference for plants of nearly the same size and age. Egg and larval counts in the field were dependent on differences of trichome density but also confounded with plant height and other characteristic gross morphological traits. Further study of plant characters affecting oviposition pre- ference and larval feeding damage should be made. Trichome density varied with environmental conditions under which the plants were grown. This variation was consistent for all lines tested. Trichome density was quantitatively inherited, and may be increased through recurrent selection. LIST OF REFERENCES 10. 11. 12. LIST OF REFERENCES Beck, 8.0. 1965. Resistance of a plant to insects. Annu. Rev. Entomol. 10:207-232. Brown, Charles L., and H.L. Shands. 1955. Factors influencing seed set of oat crosses. Agron. J. 48:173-177. Casagrande, R.A., and D.L. Haynes. 1975. Impact of pubescent wheat on population dynamics of the cereal leaf beetle. Environ. Entomol. 5:153-159. Castro, T.R., R.F. Ruppel, and M.S. Gomulinski. 1965. Natural history of the cereal leaf beetle in Michigan. Mich. Agric. Exp. Stn. Q. Bull. 47:623-653. Coffman, F.A. l955. Bagging possibly reduces success in oat crossing. Agron. J. 48:191. Cobb, D.L., B.L. Doyle, J.A. Webster, D.H. Smith, Jr., and D.K. Reis. 1978. A tractor-mounted row marker using a rotating barrel. Agron. J. 70:685-686. Cobb, D.L., and R.F. Ruppel. 1976. A self-propelled machine for mass collection of insects. U.S. Dep. Agric. Agric. Res. Serv. N. Cent. #43. Curtis, B.C., and L.I. Gray. 1957. Approach method of making crosses in small grains. Agron. J. 50:49-51. Esau, K. 1965. Plant anatomy. Wiley and Sons Inc., New York. 767 p. Everson, E.H., R.L. Gallun, J.A. Schillinger, D.H. Smith, Jr., and J.C. Craddock. 1966. Geographical distribution of resistance in Triticum to the cereal leaf beetle. Mich. State Univ. Agric. Exp. Stn. Q. Bull. 48:565-569. Everson, E.H., R.L. Gallun, and R.F. Ruppel. 1965. Host plant resistance studies of the cereal leaf beetle - Summary. Fedak, G. 1974. Effect of ligh intensity on expression of leaf pubescence. Barley Genet. Newsl. 4:24. 64 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 65 Fedak, G. 1973. Leaf pubescence in barley. Barley Genet. Newsl. 3:14. Gullun, R.L., R.F. Ruppel, and E.H. Everson. 1966. J. Econ. Entomol. 59:827-829. Gullun, R.L. l965. Yield loss in Monon wheat from cereal leaf beetle feeding. Proc. North Cent. Branch Entomol. Soc. Am. 20:65. Gallun, R.L., E.H. Everson, R.F. Ruppel, and J.C. Craddock. 1964. Cereal leaf beetle resistance studies. U.S. Dep. Agric. Agric. Res. Serv. Entomol. Res. Division Special Report W-ZOO. Hahn, S.K. 1968. Resistance of barley (Hordeum vulgare L. Emed. Lam.) to cereal leaf beetle (Oulema melanopus L.). Crop Sci. 8:461-464. Hahn, S.K. 1967. Resistance of barley to cereal leaf beetle (Oulema melanopus L.). Ph.D. Thesis, Mich. State. Univ. Hoxie, R.P., S.G. Wellso, and J.A. Webster. 1975. Cereal leaf beetle response to wheat trichome length and density. Environ. Entomol. 4(3):365-370. Johnson, H.B. 1975. Plant pubescence: an ecological per- spective. Bot. Rev. 41:233-258. Kapoor, M.L., B.D. Joshi, and A.T. Nataran. 1965. Effects of chronic gamma irradiation of epidermal hairs of some varieties of wheat and barley. Radiat. Bot. 5:265-269. Lee, Chung. 1970. The inheritance and mechanism of resistance of barley to cereal leaf beetle, Oulema melanopus L. Ph.D. Thesis, Mich. State. Univ. Leisle, D. 1974. Genetics of leaf pubescence in durum wheat. Crop.Sci. 14:173-174. Marshal, H.G. 1962. Effect of wetting and shading bags on seed set of oat crosses. Crop Sci. 2:365-366. McDaniel, M.E., and 6.0. Janke. 1973. Leaf pubescence in oats (Avena spp.). Crop Sci. 7:538-540. McDaniel, M.E., H.B. Kim, and B.R. Hathcock. 1967. Approach crossing of oats. Crop Sci. 7:538-540. Painter, R.H. 1951. Insect resistance in crop plants. The Univ. Press of Kansas. Ch. 1, 2, 3, 4. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 66 Ringlund, K., and E.H. Everson. 1967. Leaf pubescence in common wheat, Triticum aestiuum L., and resistance to the cereal leaf beetle, Oulema melanopus (L.). Crop Sci. 8:705-710 Ruppel, R.F. 1964. Control of cereal leaf beetle. Proc. North Cent. Branch Entomol. Soc. Am. 19:127-128. Sarkarung, Surapong. 1978. Influence of heredity and environ- ment of leaf pubescence in oats. Ph.D. Thesis, Univ. of Ark. Schillinger, J.A. 1969. Three laboratory techniques for screening small grains for resistance to the cereal leaf beetle. J. Econ. Entomol. 62(2):360-363. Schillinger, J.A., and R.L. Gallun. 1968. Leaf pubescence of wheat as a deterrent to the cereal leaf beetle, Oulema melanopus L. Ann. Entomol. Soc. Am. 61:900-903. Schillinger, J.A. 1967. Relationship of pubescence to resistance in wheat to cereal leaf beetle, Oulema melanopus L. Proc. North Cent. Branch Entomol. Soc. Ami 22:98. Schillinger, J.A. 1966. Larval growth as a method for screen- ing Triticum sp. for resistance to cereal leaf beetle. J. Econ. Entomol. 5:1163-1166. Schillinger, J.A., D.H. Smith, Jr., and J.C. Craddock. 1966. Laboratory screening of wheat for resistance to cereal leaf beetle. U.S. Dep. Agric. Agric. Res. Serv. Entomol. Res. Division Special Report W-217. Schillinger, J.A., R.L. Gallun, E.H. Everson, D.H. Smith, Jr., and J.C. Craddock. 1965. Cereal leaf beetle resistance studies (1964-65). U.S. Dep. Agric. Agric. Res. Serv. Entomol. Res. Division.Special Report W-207. Shade, R.E., and M.C. Wilson, 1967a. llmorphological leaf character affecting larval feeding behavior of cereal leaf beetle. Proc. North Cent. Branch Entomol. Soc. Am. 22:99. Shade, R.E., and M.C. Wilson. 1967b. Leaf-vein spacing as a factor affecting larval feeding behavior of the cereal leaf beetle, Oulema melanopus (Coleoptera:Chrysomelidae). Ann. Entomol. Soc. Am. 60:493-496. Smith, D.A., and J.A. Webster. 1974. Leaf pubescence and cereal leaf beetle resistance in Triticum and Avena species. Crop Sci. l4(2):241—243. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 67 Smith, D.H., Jr., and J.A. Webster. 1973. Resistance to cereal leaf beetle in Hope substitution lines. Proc. Fourth Ent. Wheat Genet. Symp. Smith, D.H., Jr., T. Ninan, E. Rathke, and C.E. Cross. 1971. Weight gain of cereal leaf beetle larvae on normal and induced leaf pubescence. Crop Sci. ll(5):639-64l. Steidl, R.P., J.A. Webster, and D.H. Smith, Jr. 1979. Cereal leaf beetle plant resistance: Antibiosis in an Avena sterilis introduction. Environ. Entomol. 8(8):448-450. Wallace, L.E., F.H. McNeal, and M.A. Berg. 1974. Resistance to both Oulema melanopus and Cephus cintus in pubescent leaved and solid stemmed wheat selections. J. Econ. Entomol. 67(1):105-107. Webster, J.A. 1977. The cereal leaf beetle in North America: breeding for resistance in small grains. Ann. N.Y.Acad. Sci. 287:230-237. Webster, J.A., D.H. Smith, Jr., E. Rathke, and C.E. Cress. 1975. Resistance to cereal leaf beetle in wheat: density and length of leaf-surface pubescence in four wheat lines. Crop Sci. 15(2):l99-202. Webster, J.A. 1972. Developing resistance to cereal leaf beetle, morphological basis, problems and progress. Proc. North Cent. Branch Entomol. Soc. Am. 27:98-102. Webster, J.A., and D.H. Smith, Jr. 1971. Seedlings used to evaluate resistance to the cereal leaf beetle. J. Econ. Entomol. 64(4):925-928. Webster, J.A., D.H. Smith, Jr., and J.C. Craddock. 1971. Summary of field screening for resistance to the cereal leaf beetle, Oulema melano us (L.) in the World Oat Collection. U.S. Dep. Agric. Agric. Res. Serv. Special Report W-329. Wellso, S.G. 1976. Cereal leaf beetle: feeding and oviposi- tion on winter wheat and spring oats. Environ. Entomol. 5(3):487-491. Welso, S.G., W.G. Ruesink, and S.H. Gage. 1975. Cereal leaf beetle: relationships between feedin , oviposition, mating and age. Ann. Entomol. Soc. Am. 68(41:663-668. Wellso, S.G. 1973. Cereal leaf beetle: larval feeding, orientation, development, and survival on four small-grain cultivars in the laboratory. Ann. Entomol. Soc. Am. 66(6):1201-1208. 52. 53. 54. 55. 56. 68 Wellso, S.G., R.V. Connin, and R.P. Hoxie. 1973. Oviposition and orientation of the cereal leaf beetle. Ann. Entomol. Soc. Am. 66(1):78-83. Wellso, S.G., and C.E. Cress. 1978. Intraspecific competition of the cereal leaf beetle reduced through spatial separa- tion of eggs and feeding. Environ. Entomol. 2(5):79l-792. Wilson, M.C., and R.E. Shade. 1967. The correlation of cereal leaf beetle larval populations, leaf feeding, and yields of oats. Proc. North Cent. Branch Entomol. Soc. Am. 22:99. Wilson, M.C. 1964. Host plant-cereal leaf beetle relationships. Proc. North Cent. Branch Entomol. Soc. Am. 19:124-127. U.S.D.A.—A.R.S. 1964. Watch for the cereal leaf beetle. PA-SSO.